Quilty's Script-Prediction AI Faces Reality Check After Misjudging Box Office Winners
A new AI tool claims to predict film success from scripts alone, but early tests show it confidently misread both a future flop and Oscar winner.
Last verified:
Quilty’s Pitch vs. Early Performance
Quilty, an AI startup founded by film producers Simon Horsman and Daniel Wood, entered the entertainment industry with an ambitious claim: its tool can accurately forecast a film’s success by analyzing a script. According to The Verge, the platform assigns an unproduced screenplay a score from 0 to 100, factoring narrative quality, commercial viability, audience appeal, and estimated production costs. The founders argue this insight could democratize filmmaking by helping emerging creatives refine pitches and giving financiers data-driven confidence in greenlight decisions.
The reality proved more complicated. When industry professionals tested Quilty’s predictions against actual outcomes, the tool failed spectacularly. According to The Verge, Quilty scored a script destined to become a box office flop higher than a script that would become an Oscar-winning blockbuster—precisely the kind of error that undermines confidence in its analytical claims. This mismatch between marketing and performance reveals a deeper problem: predicting film success by script alone may be inherently difficult, even for AI systems trained on industry data.
How Quilty Works (and Why It Struggles)
Unlike a purpose-built proprietary model, Quilty is assembled from multiple publicly available AI systems. Users upload a screenplay, and within minutes receive a structured report covering character profiles, narrative structure, estimated budgets, and key story beats. The service charges $50 per analysis, with volume discounts available.
This piecemeal architecture—combining existing tools rather than training a unified system—suggests the startup prioritized speed to market over technical coherence. According to The Verge, the result is “a jumbled mishmash of preexisting AI systems” with no proven ability to identify future hits. That assessment aligns with the Christy-versus-Sinners prediction failure: if your component tools disagree on what makes a film commercially viable, the final score becomes noise rather than signal.
The Human-in-the-Loop Defense
Horsman and Wood have emphasized they are not attempting to automate screenwriting or remove human gatekeepers. According to The Verge, Horsman stated the goal is to “enable human creativity” by equipping writers, producers, and studio executives with comprehensive information to make informed decisions—not to replace judgment with algorithms. Both founders acknowledged industry skepticism about AI’s impact on creative jobs and claim to have incorporated that feedback into Quilty’s design.
This positioning reflects a broader tension in AI-for-creativity tools: the pitch often requires downplaying what the technology actually can do in order to address legitimate concerns about deskilling and labor displacement. If Quilty cannot reliably score scripts, it is unclear what “better information” it provides beyond confirmation bias.
Why This Matters
Quilty’s stumble illustrates why predicting creative success remains a domain where AI tools struggle. Box office performance depends on marketing budgets, release timing, audience sentiment at the moment of release, competition, and intangible factors—none of which a script contains. Quilty’s founders may have underestimated how much the film industry’s outcome variance stems from execution and market conditions rather than story structure alone.
For producers considering AI-assisted script analysis, Quilty’s early track record suggests skepticism is warranted. Conversely, for studios and technologists building creative-AI products, the lesson is clear: high-confidence predictions on human taste are likely to fail unless the model can capture the full decision-making context, not just the artifact.
Frequently Asked Questions
What does Quilty actually do?
Quilty combines multiple off-the-shelf AI systems to analyze unproduced film scripts and generate reports scoring narrative quality, commercial viability, audience resonance, and estimated production costs on a 0–100 scale. The analysis costs $50 per script.
Why is Quilty's accuracy in question?
The platform predicted a script that became a commercial flop would outperform a script that became an Oscar-winning blockbuster, suggesting its predictive framework does not reliably identify successful films.
Are Horsman and Wood trying to automate human jobs?
No. Both founders have stated they want to 'keep humans in the loop' and provide creatives—writers, producers, financiers, studio executives—with better information for greenlight decisions, not replace human judgment.
What's Quilty's underlying technology?
Rather than a custom model, Quilty assembles existing, publicly available AI tools to perform different analytical tasks—narrative breakdown, budget estimation, character analysis, and story beat identification.